The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends installing the @membranehq/cli package from the official NPM registry. This is the standard tool provided by the vendor (membranedev) to interact with their service.
[COMMAND_EXECUTION]: The skill uses local shell commands (membrane login, membrane connect, membrane action run) to manage the integration. These commands are scoped to the functionality of the Membrane platform.
[CREDENTIALS_UNSAFE]: No hardcoded credentials or unsafe secret handling practices were found. The instructions explicitly advise against asking users for API keys, instead using a server-side OAuth flow managed by the platform.
[INDIRECT_PROMPT_INJECTION]: The skill features a search capability (membrane action list --intent "QUERY") that processes natural language strings. While this is an ingestion point for untrusted data, it is a standard search interface within a closed vendor ecosystem and poses low risk.

sesame